Amazing Job
I just wanted to thank you for such an amazing controlnet model and all the other models you have just released are excellent and on a level of their own. Those of us on the Banadoco Discord sever doing AI animation stuff have been blown away by this new model. If you want to join us just let us know otherwise thanks again!
Very happy that you like it! I am just professional in controlnet, thank you for your approval. My email is ddxin521@gmail.com, if you have any questions, you can send me an email.
I wanted to come to compliment you on your work, but somebody beat me to it.
This controlnet is banger ! Thanks
I just wanted to thank you for such an amazing controlnet model and all the other models you have just released are excellent and on a level of their own. Those of us on the Banadoco Discord sever doing AI animation stuff have been blown away by this new model. If you want to join us just let us know otherwise thanks again!
Can you introduce Banadoco to me? perhaps we can have some cooperation in the future, you can send an email to me, we can chat in detail!
I just wanted to thank you for such an amazing controlnet model and all the other models you have just released are excellent and on a level of their own. Those of us on the Banadoco Discord sever doing AI animation stuff have been blown away by this new model. If you want to join us just let us know otherwise thanks again!
Can you introduce Banadoco to me? perhaps we can have some cooperation in the future, you can send an email to me, we can chat in detail!
Here is a discord invite to the "Banodoco" group in case no one sent you one via email:
I too am very thankful for you efforts! Finally we can use CN properly w/ SDXL. If you haven't considered the one model I'd love to see you attempt to train would be the QR model similar to what the group "monster" trained for V1.5. It allows you to hide QR codes in images, but there are some other cool functions like making images w/ hidden messages that you can only see when looking at the images from a distance (The v1.5 repo: https://huggingface.co/monster-labs/control_v1p_sd15_qrcode_monster). AFAIK it has never been replicated properly for SDXL.
Additional links on the QR CN model (v1.5):
https://takin.ai/learn/generate-images-with-hidden-text-using-stable-diffusion-and-controlnet
https://old.reddit.com/r/StableDiffusion/comments/15wdoqa/been_trying_these_optical_illusions/
I have trained a QR code model for SD1. If you want to tackle it look here:
- ControlNet Mysee - Light and Dark - contains some template images for testing and helpful links
- Blobify filter in Stable Cascade - a more advanced version of the dilation and erosion filter similar to what I used for my model. interestingly it's included in Stable Cascade code but was never used afaik.
Wow! It works for Pony!